Can AI Find Its Way Into Accounts Payable?

Generative artificial intelligence (AI) is igniting a new era of innovation within the back office. No legacy process is safe. And this is particularly true for accounts payable (AP) programs, where AI, coupled with advancements in deep learning, computer vision and natural language processing (NLP), is helping drive increased efficiency, accuracy and cost savings for businesses.

“With deep learning and advancements in computer vision, AP tasks are very automatable compared to where they were 10 years back — it has made a huge difference,” Krishna Janakiraman, head of engineering at Ottimate, told PYMNTS.

According to Janakiraman, recent advancements in deep learning and computer vision have revolutionized invoice processing, making it more automatable than ever before. Tasks such as extracting structured data from images have become streamlined, reducing manual efforts and accelerating processing times.

“If you look at the whole AP automation workflow, it starts from processing invoices or documents such as statements and purchase orders, and then you are coding those documents to accounting dimensions, and ultimately approving those documents for paying,” Janakiraman explained.

This shift is enabling AP departments and finance associates to move away from manual, time-consuming processes and toward a more strategic role within their departments and organizations.

“Leaders are applying AI across the board, not just for processing invoices but also for coding invoices and approving them, all the way to payments,” Janakiraman said, noting that by applying AI end-to-end across the entire AP function, leading businesses are helping to drive more efficiency and better accuracy throughout the AP cycle, setting a new benchmark standard for the industry.

He said AI is not only automating tasks but also providing valuable insights that aid in decision-making. By analyzing vast amounts of data, AI can identify patterns and anomalies that might indicate fraud or errors, helping to enhance the security and integrity of AP processes.

AI’s Role in Transforming AP With Automation and Efficiency

While AI holds immense promise for AP automation, an effective implementation of the innovative solution is not as easy as flipping a switch that’s currently off, to on.

That’s because smaller firms may lack the resources to invest in cutting-edge technology, while larger firms may see an AP modernization exercise as something too complex or complicated relative to the “just fine” way they have always run their AP functions.

But the benefits provided by next generation AP solutions are increasingly impossible to ignore, especially when viewed against traditional, even paper-based workflows.

“There is a tremendous amount of efficiency that could be driven through using AI when it comes to automating the digitization of paper invoices,” said Janakiraman.

Still, he added that where he sees the industry heading is “more and more toward digital invoices where either you have completely digital EDI invoices, or you have invoices coming through as a digital PDF. In those cases, the lift from AI is a little less compared to paper invoices where customers are taking pictures of them to upload.”

What this means is that the need for skilled personnel to manage AI systems is becoming an emergent, and important, consideration for organizations.

Expect to See Even More Innovative Applications of AI in AP

As technology continues to evolve, we can expect to see more and more innovative applications of AI evolve the AP cycle, further transforming the function into a strategic asset for organizations.

Janakiraman said access to high-quality training data and domain expertise is paramount for building powerful AI models. Additionally, understanding customer workflows and leveraging AI talent to translate ideas into actionable solutions play pivotal roles in driving innovation in accounts payable automation.

As for where AI can play a role in streamlining AP workflows, he emphasized that AI can — and should — be applied across every industry, but that every industry is different in terms of where the most value is right now in leveraging AI.

“In the hospitality space, invoices are still very paper-based and it’s a difficult problem because the invoices are dense. The vendors use legacy printers. But if you go into the technology space or even construction, invoicing is a little bit more mature. You see far more digital invoices where AI can provide an immediate benefit by coding invoices and approving invoices,” said Janakiraman.

“If there’s no automation, then the accounting staff has to literally go through every invoice, every line item on the invoice, and map it to different dimensions,” he said.

Contrary to fears of AI replacing human jobs, Janakiraman emphasized the complementary relationship between AI and human employees. In the future, human employees will play a crucial role in teaching AI, guiding its evolution and focusing on tasks that require human judgment and creativity.

“AI allows businesses to focus on what AI cannot do — which is run and grow their business,” Janakiraman said.